This is a histogram of read lengths, in bpunits, that have a Phred-like score of 17 or better, or one error in 50 bp (the ends only are shown because of width considerations):

Distributions skewed to the right are ideal, showing longer read lengths (test fragments are a discrete length). It is likely that the sequence can extend all the way through the test fragment, if enough flows are run, so the histogram only displays a maximum size based on the length of the test fragment.